Subscriber access provided by University of Newcastle, Australia
Article
Complete Nutrient Removal Coupled to Nitrous Oxide Production as a Bioenergy Source by Denitrifying Polyphosphate Accumulating Organisms Han Gao, Miaomiao Liu, James S. Griffin, Longcheng Xu, Da Xiang, Yaniv D Scherson, Wen-Tso Liu, and George F. Wells Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.6b04896 • Publication Date (Web): 17 Feb 2017 Downloaded from http://pubs.acs.org on February 18, 2017
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 37
Environmental Science & Technology
1
Complete Nutrient Removal Coupled to Nitrous Oxide Production as a
2
Bioenergy Source by Denitrifying Polyphosphate Accumulating Organisms
3
Han Gao1, Miaomiao Liu2, James S. Griffin3, Longcheng Xu1, Da Xiang1, Yaniv D Scherson4,
4
Wen-Tso Liu2 and George F. Wells1 *
5 6
1
7
Illinois, USA
8
2
9
Champaign, Illinois, USA
Department of Civil and Environmental Engineering, Northwestern University, Evanston,
Department of Civil and Environmental Engineering, University of Illinois at Urbana-
10
3
11
Illinois, USA
12
4
13
*Corresponding author: George F. Wells, Email:
[email protected], Ph: +1-847-
14
491-3109, Fax: +1-847-491-4011.
Department of Chemical and Biological Engineering, Northwestern University, Evanston,
Anaergia, Inc., Carlsbad, California, USA
15 16
Submission as a Research Article to Environmental Science & Technology
1 ACS Paragon Plus Environment
Environmental Science & Technology
17
ABSTRACT
18
Coupled Aerobic-anoxic Nitrous Decomposition Operation (CANDO) is a promising emerging
19
bioprocess for wastewater treatment that enables direct energy recovery from nitrogen (N) in
20
three steps: (1) ammonium oxidation to nitrite; (2) denitrification of nitrite to nitrous oxide
21
(N2O); and (3) N2O conversion to N2 with energy generation. However, CANDO does not
22
currently target phosphorus (P) removal. Here, we demonstrate that denitrifying polyphosphate
23
accumulating organism (PAO) enrichment cultures are capable of catalyzing simultaneous
24
biological N and P removal coupled to N2O generation in a 2nd generation CANDO process,
25
CANDO+P. Over seven months (>300 cycles) of operation of a prototype lab-scale CANDO+P
26
sequencing batch reactor treating synthetic municipal wastewater, we observed stable and near
27
complete N removal accompanied by sustained high rate, high yield N2O production with partial
28
P removal. A substantial increase in abundance of the PAO “Candidatus Accumulibacter
29
phosphatis” was observed, from 5% of the total bacterial community in the inoculum to over
30
50% after four months. PAO enrichment was accompanied by a strong shift in the dominant
31
Accumulibacter population from Clade IIC to Clade IA, based on qPCR monitoring of
32
polyphosphate kinase 1 (ppk1) gene variants. Our work demonstrates the feasibility of
33
combining high rate, high yield N2O production for bioenergy production with combined N and
34
P removal from wastewater, and it further suggests a putative denitrifying PAO niche for
35
Accumulibacter Clade IA.
2 ACS Paragon Plus Environment
Page 2 of 37
Page 3 of 37
36
Environmental Science & Technology
TOC Art
37
3 ACS Paragon Plus Environment
Environmental Science & Technology
38
INTRODUCTION
39
Conventional biological nutrient (nitrogen (N) and phosphorus (P)) removal processes, though
40
generally reliable and efficient, are energy intensive and do not target resource recovery or
41
energy generation as a process goal1. In addition, unintended emissions of nitrous oxide (N2O)
42
via incomplete nitrification (oxidation of ammonium (NH4+) to nitrite (NO2-)) or denitrification
43
(reduction of nitrate (NO3-) or NO2- to nitrogen gas (N2) via nitric oxide (NO) and N2O) is an
44
emerging issue of concern for wastewater treatment processes targeting N removal. N2O is a
45
potent greenhouse gas with a global warming potential 310 times that of carbon dioxide (CO2)2.
46
Thus, it is usually treated as an unwanted byproduct during wastewater treatment. Paradoxically,
47
N2O is also a powerful oxidant and a potential renewable energy source. Compared with
48
stoichiometric combustion of 1 mole of methane (CH4) with oxygen (O2), roughly 30% more
49
energy is produced via combustion of 1 mole of CH4 with N2O2. Injections of N2O with oxygen
50
as a co-oxidant into a biogas-fed engine at flow rates simulating potential N2O (from waste N)
51
available from a full-scale wastewater treatment system were recently shown to increase power
52
output by 5.7–7.3%3. Due to its potential as an energy source, maximizing production, capture
53
and use of N2O for energy generation has emerged as a promising new concept for sustainable N
54
removal and energy recovery from wastewater1,4.
55
Recently, Scherson et al. introduced a novel biological N removal process, termed the
56
Coupled Aerobic-anoxic Nitrous Decomposition Operation (CANDO), to remove NH4+ from
57
high-strength wastewater (digester supernatant) and convert it to N2O for energy generation3,5.
58
Three steps are included: (1) microbial nitritation of NH4+ to NO2-; (2) microbial partial
59
denitrification of NO2- to N2O via cycling between anaerobic and anoxic conditions; and (3) N2O
60
conversion to N2 with energy production via co-combustion of CH4. Both steps 1 and 3,
4 ACS Paragon Plus Environment
Page 4 of 37
Page 5 of 37
Environmental Science & Technology
61
including measures to minimize NOx emissions during co-combustion of N2O with CH4, are well
62
documented3,6-10. In previous studies exploring high yield N2O generation from waste N via
63
microbial denitrification, a Comamonas enrichment culture and a type II methanotrophic
64
enrichment have been shown to reduce NO2- to N2O with high efficiencies using intracellular
65
storage polymers (poly-3-hydroxybutyrate, PHB) as the primary electron donor3,5,11. Pilot-scale
66
testing of CANDO for N removal from high-strength sidestreams (anaerobic digester
67
supernatant) is currently in progress.
68
Cyclic redox conditions in the 2nd step of CANDO is similar to operation of Enhanced
69
Biological Phosphorus Removal processes (EBPR) widely used in wastewater treatment plants
70
(WWTPs) for P removal12. The critical microbial functional group in EBPR processes is
71
polyphosphate accumulating organisms (PAOs). When subjected to cyclic anaerobic-aerobic
72
conditions, PAOs store endogenous carbon as polyhydroxyalkanoates (PHAs) and release P via
73
polyphosphate (PolyP) hydrolysis in the anaerobic phase, and oxidize PHAs and uptake excess P
74
to synthesis PolyP under aerobic conditions. Instead of using oxygen as the electron acceptor
75
under aerobic condition, denitrifying PAOs (DPAOs) are able to couple NO3- or NO2- reduction
76
to P uptake13,14. “Candidatus Accumulibacter phosphatis” (herein Accumulibacter) has been
77
identified as the primary PAO in most full-scale and lab-scale EBPR systems, and the ability to
78
use NO2- as well as O2 as the terminal electron acceptor has been demonstrated for selected
79
Accumulibacter lineages15,16. This suggests that it may be possible to incorporate biological P
80
removal and recovery into the CANDO process via selection for Accumulibacter-associated
81
DPAOs. The primary objective of this study was thus to clarify feasibility of combining high
82
rate, high efficiency N and P removal from wastewater with high yield N2O production as a
5 ACS Paragon Plus Environment
Environmental Science & Technology
83
novel bioenergy molecule. This 2nd generation process, termed CANDO+P, leverages a highly
84
enriched denitrifying Accumulibacter community as a microbial biocatalyst.
85 86
MATERIALS AND METHODS
87
CANDO+P Bioreactor Operation and Strategies for N2O Production
88
Since steps 1 and 3 in CANDO have been demonstrated previously6-10, we focused on step 2
89
(microbial denitrification of NO2- to N2O) in this study. A custom-built 14L (12L working
90
volume, 50% decant per cycle) sequencing batch reactor (SBR) was operated continuously for
91
seven months under cyclic anaerobic/anoxic conditions, with a short aerobic polishing step. The
92
SBR was operated with a “feast-famine” feeding strategy with synthetic municipal wastewater
93
influent under the following conditions: HRT=12h, mixing speed of 123 rpm, temperature
94
=22±1.5°C. No biomass was intentionally wasted during the operation. Biomass loss through the
95
effluent or via decay was not quantified in this study. In the “feast-famine” feeding regime,
96
biomass was alternated between chemical oxygen demand (COD) rich (anaerobic) and deplete
97
conditions (anoxic followed by a short aerobic period). COD deplete conditions were
98
accompanied by spiking with a high NO2- synthetic wastewater to mimic effluent from an
99
upstream nitritation reactor. Electron donors (COD) were switched between acetate and
100
propionate in the anaerobic phase every two cycles to enrich DPAOs17. The reactor was
101
inoculated with activated sludge from the Stickney Water Reclamation Plant (Chicago, USA).
102
The synthetic wastewater medium (without COD and N) was prepared based on Smolders et al.18
103
(see supporting information [SI] for details). Synthetic rather than real wastewater was employed
104
in this study to maintain tight control and complete knowledge of influent constituents, so as to
105
facilitate testing of feasibility (proof-of-concept) of combining high rate biological N and P
6 ACS Paragon Plus Environment
Page 6 of 37
Page 7 of 37
Environmental Science & Technology
106
removal with high yield N2O generation. Organic carbon (acetate/propionate) and NO2- pulses
107
were dosed separately to initiate anaerobic and anoxic periods, respectively. The initial COD
108
level in the anaerobic period was set by adding a pulse of stock solution (128 g/L sodium acetate
109
or sodium propionate as COD). Initial NO2- levels in the anoxic period were set by adding a
110
pulse of nitrogen stock solution (138 g/L sodium nitrite). The SBR was operated in four phases
111
to select DPAOs. Operational conditions are summarized in Table S1. In phase I (days 0-120,
112
acclimation period), both COD and NO2- concentrations were kept low. To optimize the
113
production of N2O, in phase II (days 121-158), COD and NO2- concentrations were gradually
114
increased to 100-120 mg/L as COD and 35-45 mg/L NO2- as N. Both COD and NO2-
115
concentrations were kept stable and high in phases III (days 159-181) and IV (day 182-219). In
116
phase IV, the COD/P ratio was increased to optimize P removal. In all the phases, COD/N
117
(COD/NO2- as N on a mass basis) ratio was maintained between 3-4 to facilitate maximum N2O
118
production via incomplete denitrification3,19,20.
119 120
Bioreactor Performance Monitoring
121
The SBR operation was automatically controlled by a Programmable Logic Controller (PLC).
122
The dissolved oxygen (DO) concentration and pH were continuously monitored online using a
123
LDO sensor and a differential pH probe (Hach, USA), respectively. pH was automatically
124
controlled at 8.0 ± 0.2 by addition of 0.3M hydrochloric acid and 0.3M sodium hydroxide stock
125
solutions. Liquid phase N2O in the reactor was also continuously monitored using a N2O-WW
126
sensor (Unisense, Denmark). Within-cycle SBR tests were conducted weekly to measure PO43-,
127
NO2-, volatile fatty acids (VFAs, including acetate and propionate) and intracellular storage
128
polymers (polyhydroxyalkanoates, PHAs) profiles throughout anaerobic, anoxic, and aerobic
7 ACS Paragon Plus Environment
Environmental Science & Technology
129
phases of reactor operation. Mixed liquor suspended solid (MLSS) and volatile MLSS (MLVSS)
130
were measured periodically. MLSS and MLVSS were maintained at 4000 ± 400 mg/L and 3500
131
± 350 mg/L, respectively, during stable operation.
132
PO43-, NO2-, and VFAs were assayed using an 881 Compact IC pro ion chromatograph
133
(Metrohm, Swaziland) equipped with a Supp7 column with 0.36mM Na2CO3 as eluent. MLSS
134
and MLVSS were measured according to standard methods21. PHAs (including PHB, poly-3-
135
hydroxyvalerate (PHV) and polyhydroxy-2-methylvalerate (PH2MV)) measurements were
136
adapted from Song et al. and Oehmen et al.22,23 (see SI for details).
137 138
DNA Extraction and 16S rRNA Amplicon Sequencing
139
Community structure and PAO/DPAO enrichment were monitored over time via high-
140
throughput amplicon-based DNA sequencing of 16S rRNA gene fragments. Biomass was
141
sampled periodically from the SBR reactor. Genomic DNA was extracted from 1.5 mL of reactor
142
biomass using the FastDNA SPIN Kit for Soil (MP Biomedicals, USA). The V4 region of the
143
bacterial 16S rRNA gene was amplified using universal bacterial primers 515F and 806R from
144
duplicate DNA extracts for each timepoint (see SI for details)24. Amplicon sequencing was
145
performed on an Illumina MiSeq V2 at the University of Illinois Chicago DNA Services Facility.
146
USEARCH v8.1.1861 was used to remove singletons and chimeras, and to select representative
147
OTUs based on 97% identity25,26. Phylogenetic affiliation was inferred for representative
148
sequences from each OTU using the Greengenes sequence database in the Quantitative Insights
149
Into Microbial Ecology (QIIME) platform27,28. Representative OTUs were also aligned via the
150
SILVA Incremental Aligner (SINA) to identify ‘Candidatus Competibacter phasphatis’, one of
151
the primary glycogen accumulating organisms (GAOs) that is often detected in lab-scale and
8 ACS Paragon Plus Environment
Page 8 of 37
Page 9 of 37
Environmental Science & Technology
152
full-scale reactors since ‘C. Competibacter’ is not included in the Greengenes database29,30-32.
153
All the raw sequencing data in the study can be accessed through the NCBI SRA under the
154
accession number SRP077722.
155 156
ppk1 and nirS Quantitative PCR
157
To screen for dominant Accumulibacter clades, endpoint PCR assays were applied to amplify
158
Accumulibacter-specific polyphosphate kinase 1 (ppk1) using 12 different primer sets targeting
159
11 different ppk1 Clades (including two primers for Clade IIC: Clade IIC1 and IIC2 primer)
160
(Table S2). Amplification of Accumulibacter ppk1 gene fragments was conducted in 20 µL
161
reactions on a T100 Thermal Cycler (BioRad, USA).
162
For Accumulibacter clade specific ppk1 with detectable bands after endpoint PCR
163
amplification and screening, quantitative PCR (qPCR) assays were applied to quantify the
164
abundance of these ppk1 clades on a StepOnePlus Real-Time PCR Systems (Applied Biosystems,
165
USA). qPCR assays employed 2 µL gDNA samples (100 and 1000 times dilution) as a template
166
for amplification of ppk1 gene fragments as well as amplification of bacterial universal 16S
167
rRNA genes and total Accumulibacter 16S rRNA genes. In addition to template DNA, the 20-µL
168
reactions contained SYBR Green PCR master mix (Applied Biosystems, USA) and 200 nM of
169
each primer. Details of endpoint and qPCR assays, including temperature programs, are provided
170
in Tables S2 and S3.
171
To quantify genomic potential for denitrification over time, nirS genes (encoding general
172
bacterial cytochrome cd1-type nitrite reductase) were amplified using the primers nirSCd3aF (5’-
173
AACGYSAAGGARACSGG-3’) and nirSR3cd (5’-GASTTCGGRTGSGTCTTSAYGAA-3’)33.
174
The qPCR reaction volume was 20µL with 2µL of diluted gDNA. qPCR amplification was
9 ACS Paragon Plus Environment
Environmental Science & Technology
Page 10 of 37
175
conducted in a CFX Connect thermocycler (BioRad, USA) using SYBR green master mix
176
(BioRad, USA) with the following temperature program: initial denaturation at 95°C for 10 min,
177
followed by 39 cycles of 95°C for 30s, annealing temperature of 57°C for 30s and 72°C for 30s.
178
Abundance of all functional genes (ppk1 variants and nirS) and Accumulibacter 16S rRNA gene
179
were normalized to bacterial universal 16S rRNA gene copies, and standard deviations were
180
calculated based on replicates of both dilutions for each sample.
181 182
Statistical Analyses
183
To compare N2O production efficiencies, Student’s t-tests was performed in R v3.2.434. N2O
184
production efficiencies and maximum specific N2O production rates were calculated based on
185
the liquid phase N2O concentration (mg/L as N). Gas phase N2O was ignored due to the limited
186
headspace and relatively high solubility of N2O. N2O has a high Henry’s law constant of 24
187
mM/atm (at 25°C and 0% salinity) compared with 1.3 mM/atm for oxygen, indicating that high
188
level of N2O could accumulate in the liquid phase without active aeration35. Performance
189
characteristics and statistical analyses are reported for 13 weekly full SBR cycle profiles from
190
day 127 to day 212 (phases II to IV) corresponding to the period of stable performance. Pairwise
191
Pearson correlation coefficients between total Accumulibacter 16S rRNA gene, Clades IA and
192
IIC ppk1 genes and total nirS gene were also calculated and visualized using R v3.2.4. An
193
association network was constructed to visualize the correlation between flanking OTUs and the
194
two most abundant OTUs (assigned as Accumulibacter and Zoogloea at genus level in QIIME
195
based on 16S rRNA sequencing results). OTUs with occurrence frequency less than 60% and
196
less than 100 counts were excluded from the analysis. Both a pairwise nonmetric Spearman
10 ACS Paragon Plus Environment
Page 11 of 37
Environmental Science & Technology
197
correlation matrix and Pearson correlation matrix were calculated in R v3.2.4. Network
198
visualization was constructed using circus36.
199 200
RESULTS AND DISCUSSION
201
Bioreactor Performance and N2O Production
202
To select for biomass capable of high rate incomplete denitrification for N2O generation, a
203
decoupled “feast-famine” feeding strategy was applied to the CANDO+P reactor via separate
204
delivery of external COD (electron donor) during the anaerobic phase, and NO2- (electron
205
acceptor) during the anoxic phase. NO2- removal via denitrification in the anoxic phase
206
accompanied by production of N2O (indicating incomplete denitrification) was observed
207
immediately after reactor inoculation with activated sludge. After four months of operation
208
(phase I), complete reduction of NO2- coupled to P uptake under anoxic conditions was observed.
209
By controlling the influent COD/N ratio between 3-4 and completely removing external COD
210
during the anaerobic period, N2O rather than N2 was the primary denitrification product, with
211
stable conversion efficiencies of 70-80% throughout the operational period (Figure 1a). After
212
five months of operation (during phase III and IV), as high as 35-45 mg NO2--N/L was dosed
213
during anoxic feeding to promote selection for DPAOs. pH was maintained at 8.0 ± 0.2 to limit
214
free nitrous acid (HNO2) inhibition (98%
5.1±1.6a
70-80%
a. Specific N2O production rate was calculated based on average maximum specific N2O production rate for acetate and propionate in 13 weekly cycle tests in phase III and IV.
30 ACS Paragon Plus Environment
Page 31 of 37
Environmental Science & Technology
Table 3. Comparison of specific NOx (NO3-/NO2-) and P removal rates in this and other lab-scale DPAOs studies.
Study
Electron Acceptor
Initial NO3- or NO2concentration (mg-N/L)
Specific NO3- or NO2reduction rate (mg NOx-N/gVSS/h)
Specific phosphorus uptake rate (mgP/gVSS/h)
This work
NO2-
38
6.8 a
5.7 a
Wang et al.48
NO2-
40
6.7b
9.3b
Zhang et al.49
NO2-
45
10.9c
14.4c
Zhou et al.50
NO2-
40
5.1c
5.3c
Zhou et al.50
NO3-
60
12.9c
8.4c
Wang et al.48
NO3-
34
3.6b
4.3b
a. Values were calculated based on the average of anoxic reactions of 13 weekly cycle tests in phase III and IV during first 60 mins. b. Values were calculated based on anoxic reactions during first 15 mins. c. Values were calculated based on anoxic reactions during first 30 mins.
31 ACS Paragon Plus Environment
Environmental Science & Technology
Page 32 of 37
Figure 1. Sustained high-rate reduction of nitrite and high-yield production of N2O during CANDO+P reactor operation. (a) Representative SBR profiles (phase II and III, days 127, 141, 149, and 159) demonstrating high rate NO2- reduction and N2O generation under different initial NO2- concentrations (short aeration period was not shown in the figures). Carbon source for each cycle: day 127: acetate, day 141: propionate, day 149: acetate, day 159: propionate. (b) Comparison of N2O yield with different external COD sources (acetate and propionate) based on 13 weekly full SBR cycle profiles from phases II to IV.
32 ACS Paragon Plus Environment
Page 33 of 37
Environmental Science & Technology
Figure 2. Changes across representative SBR cycles of concentrations of key compounds (COD sources (propionate/acetate), PHAs, NO2-, N2O and PO43-) in the CANDO+P SBR system using different external electron donors: (a) propionate (phase III, day 163) and (b) acetate (phase IV, day 203).
33 ACS Paragon Plus Environment
Environmental Science & Technology
Page 34 of 37
Figure 3. Overall bacterial community structure over 4 months of reactor operation (phase I and II). (a) Relative abundance based on high-throughput 16S rRNA gene sequencing of the 9 most abundant bacterial families, with the remaining detected families as well as sequences unassigned at the family level included in the category “others”. (b) Relative abundance of the 5 genera detected in the CANDO+P reactor within family Rhodocyclaceae (normalized by total universal 16S rRNA gene).
34 ACS Paragon Plus Environment
Page 35 of 37
Environmental Science & Technology
Figure 4. Circular representation of co-occurrence and co-exclusion network between flanking OTUs and the “core” CANDO+P microbiome (Accumulibacter and Zoogloea) during the startup period (phase I and II). OTUs were clustered at the 97% identity level. The inner circular diagram shows the relative abundance of 114 OTUs. Of the 20 most abundant flanking OTUs (average relative abundance > 0.8%), 11 that significantly co-occurred or co-excluded with Accumulibacter or Zoogloea are numbered and labeled based on the lowest level of taxonomic assignment in QIIME using the Greengenes reference database. The outer circular diagram shows the relative abundance of different phyla (phyla with relative abundance < 1% or contained